SolvEdge App

PoseTrack is an Android application that uses real-time camera input and machine learning to track elbow movement for physiotherapy exercises. The app leverages MediaPipe Pose, CameraX, and a custom Node.js + PostgreSQL backend to log session data with angle-specific image snapshots and time tracking.

Key Features

• Elbow Pose Detection (left/right) using MediaPipe's heavy model
• Live Angle Tracking with overlay visualization
• Automatic Snapshot Capture at 135°, 90°, and 45°
• Time Taken Logging for each target angle
• Video Recording of each session
• Session Summary Screen with video and snapshot cards
• Cloudinary Upload for snapshot images
• Backend Integration (Node.js + PostgreSQL on Render)

---

How It Works

1. Launch the app and tap Start Session.
2. Choose which joints (elbows) you want to track.
3. The app detects pose in real-time, captures snapshot when a target angle is reached.
4. Once done, click End Session to stop recording and upload logs.
5. View session data (video, joint angle, time, snapshot) on a summary screen.

---

🛠 Tech Stack

Android (Frontend)

• Kotlin, CameraX, MediaPipe PoseLandmarker
• Custom overlay (PoseOverlayView)
• Session logging via OkHttp
• UI built with ConstraintLayout, CardView, and RecyclerView

Backend

• Node.js with Express.js
• PostgreSQL (via Render.com)
• Image upload to Cloudinary
• REST APIs: /log-snapshot, /session-logs?session_id=

---

🚀 Project Setup

Prerequisites

• Android Studio Giraffe or later
• Android device with camera access
• Backend deployed to Render (see Backend Repo)

Clone & Run

git clone https://github.com/your-username/pose-track-android.git